Forty Billion Nodes Under The Tree

The Newsletter of the British Othello Federation

July 1993

© •• ©

Black to play, but White wins 20-16.

See page 6 for further details.

Othello is manufactured and marketed by Peter Pan Playthings, Peterborough.The British Othello Federation is an independent body. Annual subscription forBritish residents costs £5 for the first year’s membership (including a copy ofthe instructional book Othello: Brief & Basic) and £3 thereafter. Ten yearsmembership is available for £25. An overseas subscription costs £5 per year,or £45 for ten years. Cheques or postal orders payable to the British OthelloFederation should be sent to David Haigh. The price of Othello: Brief & Basicfor existing members is £5.

Contents.Proposed Rule Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Graham Brightwell . . . 5Amenor Wins World 6× 6 Championships! . . . . . . . . . . . . . . . Joel Feinstein . . . 6The 1993 Regional Tournaments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Other tournament reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Endgames from the Nationals . . . . . . . . . . . . . . . . . . . . . . .Graham Brightwell . . . 16Perpendicular Worlds Discovered . . . . . . . . . . . . . . . . . . . . . . . .David Shaman . . . 20A Really Hard and REALLY Pointless Othello Puzzle . .Aubrey de Grey . . . 26Advertising Thor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Joel Feinstein . . . 28Bat out of Hell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Graham Brightwell . . . 29The Art of Sacrificing Four Corners . . . . . . . . . . . . . . . . . . . . . . Joel Feinstein . . . 34The 1993 Cambridge Open . . . . . . . . . . . . . . . . . . . . . . . . . . . Karsten Switness . . . 37Two Bees with One Stone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marc Tastet . . . 38Syncopated Cerebrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sid Cox . . . 43Correspondence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . edited by Graham . . . 45To Rate or Not to Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David Haigh . . . 46The Rating List . . . . . . . . . . . . . . . . . . . . . . . . . . . .maintained by David Haigh . . . 48

Addresses of Officers.

Chairman and Newsletter Editor: Graham Brightwell, 12A Glenfield Road, Bal-ham, London SW12 0HG. Tel. 081-675-8873 or 071-955-7624 (Work).Secretary: David Haigh, 62, Romsey Road, Winchester, Hants., SO22 5PH. Tel.0962-853826.Treasurer and Deputy Editor: Peter Bhagat, 1 Parker Street, Cambridge CB11JL. Tel. 0223-62323.Publicity Officer: Mark Wormley, Flat 45, Rowntree Wharf, Navigation Rd.,York. YO1 2XA. Tel. 0757-618181 x3836 (Work).

The Editor would like to take advantage of this space to thank all those who sentin articles so promptly. One or two items had to be held over to next time, butthat’s the way the Editor likes it!

2

Forthcoming Events.This year’s National Final will be held, as previously announced, on the weekendof August 21-22. However, not as previously announced, it will be held in thefair city of Sheffield. Many thanks to Robert Stanton for finding us a venue atshort notice.

The venue is the Charnwood Hotel, 10 Sharrow Lane, Sheffield, just offLondon Road. The schedule is the same as in previous years, so the AGM willstart at 1:00p.m. on Saturday 21st August, with the National Final starting at3:30. Play will continue on Sunday 22nd, starting at 9:00a.m. As is becomingtraditional, the climax of the event will be a Grand Final beginning at about5:00p.m. The usual Challengers’ event, open to all those not qualified for theNationals, will be held alongside on the Sunday. Accomodation is available atthe Charnwood hotel at a reduced rate of £25 per person per night, but Robertwill also investigate cheaper alternatives. The phone number of the hotel is 0742589411. Robert’s phone number is 0742 855242 and he claims to be nearly alwaysthere just after 6:00p.m. His work number is 0742 824307.

Travel details will be available to the qualifiers and other interested partiesat a later date.

The 1993 World Othello Championships will be held in Kensington, London,over the weekend of November 5-7. On the first two days (Friday and Saturday),there will be a Swiss-style tournament to decide the World Team Championshipand to find the four semi-finalists. The semi-finals, final, and Victory Dinner takeplace on the Sunday.

We are still looking for more volunteers to act as table referees: thanks toeveryone who has replied so far. Being a referee involves taking the score ofa game, checking for flipping errors, watching for time defaults, and generallywatching some top quality games. Meals will be provided. Anyone who thinksthey might be interested should contact Graham Brightwell after August 9th.Don’t worry that you’re not a good enough player; all we need are people whohave played a little. Further details will be available from Graham nearer thetime, hopefully in the next issue of the Bulletin.

The remaining European Grand Prix tournaments are as follows.Brussels — July 31-August 1; Paris — August 28-29. All the usual recommenda-tions hold. Go. The Paris tournament this year will feature three top Japaneseplayers: Takeshi Murakami, Mazaki Takizawa and Hideki Kitajima, not to men-tion the four top Russians.

3

Secretarial Snippets by David Haigh.

Some members have been perplexed to receive a membership renewal request amere 6 months after shelling out £5 for their first year’s membership. Here is howthis can happen. Remember that a year’s subscription pays for two newsletters.Say you join in early July, and are sent the July newsletter as the first newsletterpaid for by your subscription. Your second newsletter will be sent to you inJanuary, and since it is the last one you have paid for you will find a renewal slipwith it. (In order to save postage costs we do not send out separate reminders.)You don’t have to renew your membership straight away; so long as you do sobefore July, you will get the next newsletter. But we hope that most people willrenew immediately, because it is so easy to forget to do so.

About Last Issue’s Cover by Karsten Sotherwitness.

It was written in the last issue that a complete British ‘B’ team was enough towin the team World Championships. Maybe it should be stressed that the Frenchteam in Barcelona—Tastet, Juhem and Penloup—was also a complete ‘B’ team:the French ‘A’ team—Ralle, Piau and Puget—won the title in Stockholm in 1990.

Furthermore, someone pointed out to me that this French ‘B’ team hadactually also won the team Championship in Barcelona! If you consider the finalresults, Feinstein is equal with Penloup, Handel is equal with Juhem, and Tastetis ahead of Shaman, so France is ahead of Britain, isn’t it? [No. – Ed.]

Even better, as David Shaman is moving to France and will be eligible toplay in the French team from 1994, even a French ‘C’ team composed of Shaman,Nicolet, and Caspard should be enough to win the team title in 1994. If you’renot convinced, have a glance at the results of this year’s Cambridge Open.

Notation.

a1 b1 c1 d e f g h1a2a34567 X C8 C h8

The board is split into eight columns and eightrows. Each column is labelled with a letter, from‘a’ for the left-hand column to ‘h’ for the right-hand column. Rows are numbered from ‘1’ forthe top row to ‘8’ for the bottom. This is theopposite convention to that used in chess. Thusthe top left corner is called ‘a1’, and the bottomright is ‘h8’.

A square such as b2, one in diagonally froma corner, is known as an ‘X-square’, and a squaresuch as h7 adjacent to a corner is a ‘C-square’.

Compass directions are sometimes used when describing regions of the board,so for instance the area of the board near to h1 is called the North-East corner.

4

Proposed Rule Changes by Graham Brightwell.

Your faithful committee met twice this year, once in December and once in June.In many ways, it’s been a quiet year, but the Committee did get round to consid-ering some minor changes in the Rules governing qualification from the Regionalsto the Nationals. These will be put to the AGM, to be held on Saturday August21st, just before the Nationals.

(1) Phil Marson noted that, in the past, players qualifying for the Nationalshad declined to take up their place in the tournament. The effect is that thereare fewer players in the event than there might be, while players who are keen toplay are excluded. Phil proposes that, in the event of a player (X) declining theinvitation to play in the Nationals, the next highest player, not already qualified,in the tournament where X originally qualified, would be invited to play. Thecommittee were a little concerned about a few possible consequences of this,and felt that it would only be workable if X declined, in writing, the formalinvitation to play in the Nationals (which goes out after all the Regionals havebeen completed). Apart from this, the committee felt it was workable, and byand large had no strong feelings for or against the proposal. A formal motion willbe put before the AGM, with no committee recommendation. According to ourshiny new constitution, a simple majority at the AGM will suffice to implementthe rule change. Of course, the rule will not operate in 1993.

(2) On a couple of occasions in the past two seasons, a player has qualifiedfrom a Regional by virtue of scoring 0 points, but being among the top threenon-qualifiers present. It seems that players are a little embarrassed by this,and would rather not have qualified under such circumstances. It was thereforeproposed that, in order to qualify for the Nationals, a player must (a) be amongthe top three finishers at a Regional, among those players not already qualified,and (b) have scored at least half a point, not counting byes. I point out that thecommittee has the right to suspend the rules and allow others to qualify if thecircumstances are unusual – this has happened once or twice in the past. I canenvisage a case where only one player turns up at a Regional, hoping to qualify. . . . Again, the committee didn’t have any very strong feelings for or against thisproposal, and will not be making a recommendation to the AGM.

It’s daft scenario time. At a Regional, there are four players not alreadyqualified, two of whom do well. Player three scores one point by beating Bye.Player four scores half a point by drawing with A.N.Human. Which of them qual-ifies? If we adopt proposal (2) above, the answer is that neither qualifies. Playerthree falls foul of requirement (b), while player four hasn’t satisifed condition (a).This anomaly was pointed out by Imre Leader. In practice, of course, playersthree and four will play each other, and this won’t arise. But, when it does . . . .

5

Amenor Wins World 6x6 Championships! by Joel Feinstein.

A great deal of excitement preceded the long awaited three game match betweenAmenor and Urakam, two players known to play almost perfectly. The matchproduced three games of exceptional quality.

Computer analysis reveals that Amenor made no errors, while Urakam’smove 2 in game 1 and move 5 in game 2 cost him one disc each. In game three,play was perfect from move 2 onwards.

b2 c2 d2 e2 f2 g2b3b4 © •b5 • ©b6b7

Notation.In the following notes, the columns are labelled b to

g, and the rows are labelled 2 to 7, so that the moves instandard openings have their usual names.

Colours are red and white, with red playing first.Interest in 6 × 6 Othello is on the increase, because

Peter Pan is planning to market the game, targetted atthe 5-8 age range. It seems that the game is playable,and incorporates many, though not all, of the interesting

features of the full game.

22 28 5 26 27 2123 11 1 2 14 2913 8 © • 3 615 4 • © 9 3124 16 7 10 30 3219 25 12 18 17 20

Game 1

Game 1. Red: Amenor 17, White: Urakam 19.The first surprise of the match came with Urakam’s

choice of 2.e3, the parallel opening, commonly held tobe inferior. At move 5, the usual move in eight by eightOthello is at f3, but this is an x-square, so instead Urakaminitiates play on the North edge. White now has littlechoice but to grab the East edge. Many onlookers nowliked the look of the move 7.f5, but computer analysisreveals that after the optimal moves 8.g5 c4 b4, Red has

no good way to break the long flat White wall, and White wins 20-16. Amenor’schoice of 7.d6 was more accurate. Play is then fairly natural until Amenor’s bravediagonal grab at move 11. Urakam played the waiting move 12.d7, but after 13.b4he was forced into an x-square of his own. 14.e2 would lose to 15.f6, while 14.f2loses to the Stoner trap 15.e2. In the game, after 16.c6, Amenor allowed theStoner trap on the South edge but ensured his own access to g2. The playersthen captured the four corners on consecutive moves. Unfortunately for Amenor,White’s threats in the North are too strong. In the end Urakam was able to playboth e2 and c2, and Amenor did not even gain parity, as the two regions in theEast were linked. These regions did, however, play quite well for Amenor, andhe was able to keep the loss to 17-19.

6

21 28 6 9 10 2911 16 1 4 27 3012 7 © • 5 248 2 • © 14 23

13 25 3 15 17 2032 26 31 22 19 18

Game 2.

Game 2. Red: Urakam 15, White: Amenor 21.In game two, the players played down the main line of

the Rose opening up to move 5, but Amenor avoided thex-square 6.c6 in favour of the aggressive sequence 6.d2 c4b5 e2 f2. Now 11.c3 looks natural, but Urakam chose notto sacrifice the b2 corner. At move 14, g4 looks strong,but Red could then play the sacrifice at c3. This turnsout to be a 19-17 White win, but Amenor wanted more!After 14.f5 e6, Amenor grabbed the diagonal with 16.c3.

© © ©• © © ©• © © © •• © • © ©• • •After 16.c3.

This diagonal grab is extremely strong, as it preventsRed’s f3, and threatens c6. If Red tries 17.g5, he nevergets on the diagonal after 18.e7! (White wins 26-10).

In desperation, Urakam tried 17.f6!, following the“principle of the opposite x-square”. This huge sacrificeguarantees Red’s access to b2. But Amenor cleverly de-nied Urakam access to c2, preventing him from workingalong the North edge until it was too late. Note that 22.g5loses to 23.g4, gaining access to c2. After 25.c6, b7 and

c7 are equally good, but c7 is more spectacular, with its follow-up 28.c2.Ahead on disc-count, Amenor chose white for the deciding game.

28 9 8 7 31 3222 14 1 4 17 1627 13 © • 6 1519 2 • © 5 1224 23 3 10 26 2921 18 20 11 25 30

Game 3.

Game 3. Red: Urakam 16, White: Amenor 20.At move 5, Urakam chose a different variant of the

Rose opening, and play proceeded down the most com-mon line until move 9, when Urakam grabbed the Northedge. After 10.e6 it looked as if he was in trouble, but thedramatic centre grab 11.e7 turned out to be quite strong.Both players were rather short of moves, and it was clearthat an x-square was imminent. Amenor took the initia-tive with 14.c3! As often, this diagonal is very hard to

cut. Urakam did not even try, and at move 17 he sacrificed at f3.

• • •© • • • ©© © • • ©© • © • ©• ••

After 17.f3

The position was now extremely tight. Neither sidecould cut the other’s diagonal in a satisfactory way. AfterRed’s waiting move to a5, Amenor played the surprisng20.d7. After 21.b7 b3, the point is not that Amenor wason the diagonal, but that Urakam could not cut the otherdiagonal, and also that Amenor was certain to keep par-ity in at least three of the four regions. Note that thistime the opposite x-square 25.f6 would be no help after26.g7. Amenor had enough for a 20-16 win, giving him

the championship. Congratulations to him.

7

In case you are wondering, the characters Urakam and Amenor are fictional, butthe computer analysis is genuine. The six by six version of my program MODOT,running on the Nottingham Mathematics department’s Silicon Graphics crimsonchip, took about two weeks cpu time to show that the sequence in game threeabove is optimal. It had to look at about forty billion positions. Previously,the win, loss or draw calculation had taken about 1.5 weeks and thirty billionnodes to show that White was winning at move 2. The program used short mid-game searches to order the moves for its endgame calculation to look at. Usuallythis was sensible, but some of the corner sacrifices looked at were daft. On theother hand, the optimal lines do include a lot of early x-squares. I would notrecommend an x-square at move 2, though! The diagonal opening is an easy winfor Red.

The following opening tree is enough to guarantee that White is winningafter three moves each. I do not claim that White’s moves are optimal. Also, thewin after d3 c5 b6 e3 d6 c4 b5 is very hard to find! (This line is not optimal forWhite).

d3 c5 b6 e3 f5 e6 f5 f4f4 d2 f3 e2d6 c4 e6 d2 c2 b2f2 d2 c3 e3f6 d2 c4 b5f3 d2 b5 f4

c6 e3 b5 d2 c6 f4c4 b5 f6 d2 c4 b5f5 d2 f6 d2 c4 b5f3 f4 c3 f5

d6 e3 b4 d2 c6 f5f4 d2 c2 b2b5 d2 b5 f4

Further information: if after 2.c5 Red tries 3.e6, White can win by morethan 4 by playing the ganglion, 4.d2. But if White plays 4.f5 instead, Red mustplay the Italian 5.c4, and it’s a draw: perfect play is then 6.e3 f4 g5 g4 g3 f6! d6b5 g7 e7 c6! b7 d2 d7 f7! g6 f3 g2 c7 f2 e2 c3 b6 c2 b2 b3 b4 18-18.

I have credited empty squares to the winner, as in tournaments. It is possiblethat there is a line where Red can hold the loss to 16-17. I do not intend to useanother 2 weeks cpu time to find this out!

8

The 1993 Regional TournamentsThe results of all the Regionals are as below, followed by some reports. Thanks toall those who sent us reports: I hope none got lost this year. While we’re on thesubject, I can exclusively reveal that the report of the 1992 Edinburgh Regionalhas now been found, but sadly not in time for it to make this newsletter. Weplan to feature this in the next issue of the popular British Othello Bulletin.

Ties are officially broken, and I believe that I have the finishing order correctin all cases. There were some errors in this respect in the results as in the Bulletin:I would like to apologise for the distress that this has apparently caused in someplaces. Joel Feinstein and Neil Stephenson had already qualified by winning theNationals and the Challengers’ respectively last year. So we have 29 qualifiers sofar, with a possible three more from the Doncaster Regional.

London – 13/3 Pts Wellingborough – 3/4 Pts1. Peter Bhagat (Q) 5/6 1. David Shaman (Q) 7/72. Joel Feinstein 5 2. Graham Brightwell (Q) 63. Michael Trent (Q) 4 3. Aubrey de Grey (Q) 54. Iain Barrass (Q) 4 4. John Lysons 45. Ian Turner 3 5. Iain Barrass 46. Phil Marson 3 6. Joel Feinstein 47. Mark Atkinson 3 7. Martin Fancy 48. Neil Cuthbertson 3 8. Guy Plowman 39. John Bass 3 9. Phil Marson 310. Aiden O’Reilly (IRE) 2 10. Roy Arnold 311. Liam Stephens 1 11. Margaret Plowman 312. Steven Verhaegen (IRE) 0 12. Adelaide Carpenter 2

13. Nigel Barthorpe 1

Eastbourne – 17/4 Pts Worksop – 24/4 Pts1. Guy Plowman (Q) 7/7 1. Joel Feinstein 7/72. Phil Marson (Q) 6 2. Iain Barrass (Q) 53. Ian Turner (Q) 5 3. Iain Forsyth 54. Graham Chappell 4 4. Phil Marson 45. Roy Arnold 4 5. David Haigh (Q) 36. Alison Turner 3 6. Mark Wormley (Q) 27. Richard Brend 3 7. Colin Hands 28. Ashley Hammond 2 8. Eileen Forsyth 09. Rodney Hammond 110. Gareth Taplin 0

9

Nottingham – 8/5 Pts Winchester – 22/5 Pts1. Mike Handel (Q) 6/6 1. Imre Leader (Q) 7/72. Joel Feinstein 4 2. Guy Plowman 53. Graham Brightwell 4 3. Aubrey de Grey 54. Guy Plowman 4 4. Ian Turner 45. Iain Barrass 4 5. Mark Atkinson (Q) 36. Jeremy Das (Q) 3.5 6. Roy Arnold (Q) 27. Ken Stephenson (Q) 3 7. Adelaide Carpenter 18. Mark Wormley 3 8. Phil Marson 19. Roy Arnold 310. Daniel Olivares 3 Manchester – 5/6 Pts11. Phil Marson 2.5 1. Jeremy Das 5/612. Colin Hands 2 2. Mark Wormley 413. Iain Forsyth 2 3. Ken Stephenson 414. David Haigh 2 4. Phil Marson 415. John Simpson 2 5. Jim Brewer (Q) 416. Eileen Forsyth 0 6. Phil Brewer (Q) 3

7. Iain Barrass 3Cambridge – 19/6 Pts 8. Roy Arnold 2

1. Imre Leader 5.5/7 9. Colin Hands (Q) 12. Garry Edmead (Q) 5.53. Guy Plowman 5 Portsmouth – 12/6 Pts4. Graham Brightwell 5 1. Graham Brightwell 7/75. Aubrey de Grey 4.5 2. Ian Turner 66. Peter Bhagat 4 3. Aubrey de Grey 57. Phil Marson 4 4. Phil Marson 48. Ian Turner 4 5. Ali Turner (Q) 39. David Haigh 4 6. Roy Arnold 210. Iain Barrass 3.5 7. Neil Cuthbertson (Q) 111. Jeremy Das 3 8. Adelaide Carpenter (Q) 012. Roy Arnold 313. Matthew Selby (Q) 3 The remaining Regional14. Jeremy Rickard (Q) 3 will be held on July 24th,15. Adelaide Carpenter 3 in Doncaster. Will Phil16. Graham Chappell 2 Marson play all 10 events?17. Gareth Thomas 1 Will someone break 30018. Simon Nickson 0/3 rated games?

10

London Regional by Graham Brightwell.

Oldcomer Mark Atkinson returned: he now lives and works in London. Twotrundled over from the Emerald Isle, wherever that is. Aiden O’Reilly was disap-pointing, but Steven Verhaegen did quite well considering he’d only been playingfor six weeks. He would have won his first round game against Liam Stephensif he’d flipped all the available discs at move 59. A more impressive debut washad by Michael Trent: in fact, has there been a more impressive debut in recentyears? He is apparently one of Britain’s top Shogi players, which undoubtedlyhelps, and he has played a number of correspondence games. His first humanOTB victim may well have been, er, me, when he turned up at the Grotto Clubthe month before. This tournament saw him start with a 63-1 win over NeilCuthbertson, then lose steadily to Pete, then lose to Aiden, beat somebody, beatIan Turner (!) and then beat Phil Marson in a key last round game. Basically thewinner took the final qualifying place: Phil was winning going into the ending,and I think both 57 and 59 were half-game-losing errors. 33-31 to Michael. Atthe top of the tournament, Joel lost to Phil Marson, but beat Pete. That’s it. Irefereed smoothly and ably.

Wellingborough Regional by Bye.

I wish to register a complaint. I turned up on time for the abovementionedregional, along with 12 other players, five Othello sets, a so-called Reversi setfrom Russia, and a quorum of clocks. Play commenced with me paired againstthe absent Graham Brightwell. I was hardly able to contain my excitement asI was poised to gain 32 rating points. But then I was shattered to learn that,in his absence, he had been awarded a point, whereas I, through absolutely nofault of my own, scored 0. Now, I am used to defeat, even against the weakest ofplayers, but I feel that the organisers here inflicted a gross humiliation on me. Inthe rest of the tournament, I laboured under the extra pressure, and sadly failedto score. Then, in the last round, fresh hope arose. I found myself paired againstJoel Feinstein, whom I had just seen dashing off to catch a train. But once againmy hard-earned point was cruelly stripped from me by harsh organiser MargaretPlowman. The result was that I finished an undeserved last. This experienceruined what should have been my greatest ever sporting day, when the racehorseI own romped in first in the Grand National.

I didn’t notice much about the rest of the tournament. David Shaman seemedto be playing well. I saw Aubrey de Grey beat Joel Feinstein. The rumouredreturn of Garry Edmead turned out to be just that. David, Graham and Aubreyqualified. Margaret Plowman refereed smoothly and ably, apart from her crueltreatment of my good self.

11

[The Editor wishes to apologise to Margaret Plowman, and indeed to anybodyelse, for any offence this piece might have caused. The Editor also wishes to stressthat it wasn’t Margaret’s fault that the sets didn’t arrive: he can’t rememberwho’s fault it was, fortunately, but it was someone in Cambridge.]

Eastbourne Regional by Rodney Hammond.

Another fine day at the seaside as Guy Plowman, Richard Brend, Roy Arnoldand Phil Marson crossed the North/South divide to join Ian and Alison Turner,Rodney Hammond and his son Ashley, newcomer Gareth Taplin and, fresh outof retirement, Graham Chappell.

The first round went to form with wins for Guy, Ian, Phil and Roy. Grahamimpressed by losing just 31-33 to Ian Turner in his first competitive game for fiveyears. Rounds 2 and 3 progressed, with Guy and Phil the only unbeaten players.In round 4 they met, with Guy winning 36-28 after a close game.

No surprise results meant that Guy (7 wins), Phil (6) and Ian (5) were thethree qualifiers.

Worksop Regional by Joel Feinstein.

This was run very efficiently by Roy Arnold. There were 8 players, so we playeda seven round round-robin tournament. I should have lost to David Haigh, butthen he should have lost to Colin Hands. The three qualifiers were Iain Forsyth,David Haigh and Mark Wormley (on tie-break). The two Iain’s did well. PhilMarson recovered well from losing his first three games.

Nottingham Regional by Phil Marson.

The 1993 Nottingham Regional attracted a record 15 players. OK!! Last year’scompetition was the first regional to be held in Nottingham so there wasn’t muchto beat; nevertheless, 15 was a great turnout and made for an interesting tour-nament.

To avoid a bye, I decided to play. Considering the pressure involved inorganising a tournament, doing the pairings for each round and playing, I shallcertainly never do that again. [Editor’s Note: the fact that Phil managed therebyto play in all the Regionals this year is irrelevant.]

My thanks to all those who made the effort to travel such great distancesto be with us. Players arrived from London, Winchester, Cambridge (of course),Newton Aycliffe and Doncaster. However, first prize (if there had been one) mustgo to Daniel Olivares, who made his way from Buenos Aires especially for thistournament! (Other, more cynical, people may say he was visiting relatives inEurope; that he had heard about the competition whilst staying in France and,

12

since he would be visiting England and staying nearby, decided he would like toplay; but I’m sure you know whom to believe.) Anyway, Daniel told me thereare about four players in the whole of Argentina so he doesn’t get too muchpractice. However, he acquitted himself well and appears to be a competentplayer. [Editor’s Note: someone is exaggerating. I am often sent results ofArgentine tournaments, and there must be at least eight players.]

There were plenty of non-qualifiers competing for the three National places,but in the end they went to Mike Handel, Jeremy Das and Ken Stephenson.Congratulations to Mike, who won all of his games against a strong field, andtook the first prize of a cuddly dog in a fireman’s helmet, and thanks to all thosewho took part.

Winchester Regional by David Haigh.

For the second year running eight players attended the Winchester Regional. Thedifference this year was that more than three of them had not yet qualified, so atlast the tournament had some point to it, apart from the fun of playing Othello.Those seeking qualification were Roy Arnold, Mark Atkinson, Adelaide Carpenterand Imre Leader. Supporting them in this endeavour were Aubrey de Grey, PhilMarson, Guy Plowman and Ian Turner.

The surprises in the pre-lunch play were Adelaide’s 33-31 win against IanTurner in round 1, and Aubrey very nearly beating Imre in round 3. Two morediscs would have been enough. Also in this round Adelaide did well to comewithin three discs of beating Phil. At the lunch break the scores were: Imre 3/3,Aubrey, Guy and Ian 2, Adelaide and Phil 1, Mark and Roy 0.

Imre was delighted to learn that lunch would be taken at the Jolly Farmer,the pub which had given him much gastronomic delight on previous occasions.Imagine my concern when I saw from the huge array of pump handles and casksbehind the bar that the place had clearly been converted from a dessert-eater’sto a beer-drinker’s pub. However, this could not have been too upsetting for Imrebecause he went on to win all his remaining four games and the tournament.

Roy provided two post-lunch surprises with his 39-25 defeat of Phil in round4 and in the 7th and last round by only just losing to Mark in the third 31-33game of the day.

Using Sonneborn-Berger to split the ties, the results were: Imre 7/7, Guy 5(12), Aubrey 5 (11), Ian 4, Mark 3, Roy 2, Adelaide 1 (4), Phil 1 (1). Imre, Markand Roy therefore qualified for the Final.

[Sadly, we have no report of the Manchester Regional, probably because we forgotto ask for one. Apologies to all those who did well, and expected to see their featsreported. Rumours filtered through that Jim Brewer was in this category, andJeremy Das’s victory should not go unremarked.]

13

Portsmouth Regional by Ian Turner.

Preparation for the tournament started on the Friday night: I drove aroundPortsmouth picking up first chess clocks, kindly loaned by Portsmouth ChessClub, and then Aubrey de Grey, Adelaide Carpenter and Roy Arnold, who hadcontrived to arrive in the same place at the same time. The next morning we werejoined at the venue by Neil Cuthbertson, Phil Marson and Graham Brightwell.

Starting out as favourite, any ideas that Graham had about this being an easytournament were quickly dispelled when he was pushed almost to the endgame byAlison, who started the tournament as the lowest rated player. Wins by Aubrey,Phil and myself set the pattern of the day’s play.

By round four we began to see a tournament of two halves, with Grahamand myself (4/4), and Aubrey and Phil (3/4) being well clear of the field. Inround five, Graham beat Phil, and then came round six with the critical gamebetween Graham and myself. Graham won 63-1, but with Aubrey beating Phil,and Alison jubilant after beating Roy for the first time, the tournament was readyfor a nail biting final round, as Graham was just one point ahead of Aubrey andmyself, and his final game was against Aubrey.

I finished the tournament with a win and when I went over to watch Gra-ham and Aubrey it looked like a three-way tie was the most likely result. How-ever, under the watchful gaze of the rest of the competitors, Graham “Houdini”Brightwell, ably assisted by Aubrey, managed not only to escape but even torecord a 50-14 win, taking the tournament with an unbeaten record.

No-one sent us a report of the Cambridge Regional either. Obviously the inexpe-rienced organiser couldn’t be expected to realise that we’d like one. Fortunatelythe Editor went to that tournament himself, and has a dim recollection of beingpaired against Imre Leader in Round 1 (losing), losing to Phil Marson in Round 2,but then bouncing back to finish with a crushing victory over the inexperiencedorganiser (Britain’s eighth best player).

Garry Edmead’s comeback featured an ignominious first-round loss againstAubrey de Grey, then later a pulsating draw with Imre Leader. Curiously, thiswas one of the few Regionals not preceded by rumours of an Edmead comeback.Leader lost to Plowman, but won the tournament anyway. Further down thegratifyingly large field, Gareth Thomas returned after an even longer absencethan Garry’s.

Thanks are due to the aforementioned organiser, and indeed to all those whoorganised or helped with any of the Regionals this year. Thanks are also dueto those who played: the turnouts at Nottingham and Cambridge in particularwere higher than we’ve seen for a couple of years, and this does make for a moreentertaining tournament for all concerned.

14

I think the next two reports speak for themselves – Ed.

Winchester Beginners’ Tournament by David Haigh.

Despite indications from several new members that they intended to participatein the first Winchester Beginners’ tournament, none of them managed to makeit and the only potential participants (apart from me) were Ali Turner and anunrated packet of biscuits.

Ali made a start at demolishing the packet of biscuits, but this did notmanage to respond and lost on time (thereby saving itself from being completelywiped-out). Ali was therefore declared the winner of the tournament, and winsa year’s membership of the BOF.

The expert on hand was Ian Turner, and he too gets a year’s membershipfor his trouble, and for the excellent advice he gave Ali on the opening techniqueto use against the biscuits.

Nottingham Junior TournamentPhil Marson had a bright idea for the Nottingham Junior Tournament: get anarticle in the local paper, and offer a year’s free membership of the Federation forthe first 16 people to reply. Dismally, the paper reported this as “There will bea tournament sometime in the next year”, with no mention of the exciting offer.

On the day, Phil and “on-hand expert” Joel Feinstein were joined by ColinHands and his brother Jonathan, and the ubiquitous Roy Arnold. At least thatwas enough to hold a sensible round-robin tournament, which they duly did. Philreports that “The results were irrelevant.” Joel reports that Joel won it with 4/4,Phil scored 3, Roy 2, Colin 1 and Jonathan 0. Perhaps Phil was right. Our legaldepartment is still sorting out who, if anyone, wins a year’s free membership.

* * * * * * * * * *

Mini-tournament, February 6th, Doncaster, by Eileen Forsyth.

Joel came, he saw, he won.Eight of us played in the four-round tournament, with Sue Barrass doing the

pairings. Most of our usual Thursday players were there, with the exception ofMaurice Kent and John Beacock, who were playing chess.

Round 1 was uneventful. Round 2 featured Phil Marson beating Iain Barrass34-30, and Mark Wormley producing a healthy 28-36 loss against Joel. This scorewas echoed in the final round by Iain Barrass, again against Joel.

Final results. 1. Joel Feinstein 4, 2. Phil Marson 3, 3. Mark Wormley 2 (10),4. Iain Barrass 2 (9), 5. Roy Arnold 2 (7), 6. Iain Forsyth 2 (6), 7. Colin Hands 1,8. Eileen Forsyth 0.

15

Endgames from the Nationals by Graham Brightwell.

A welcome innovation at last year’s National Championships was that the tran-scripts were collected at the end, and packaged together. Participants at the Na-tionals should receive the transcript collection with this newsletter, if they haven’talready. Credit for this goes to David Shaman, who typed all the games in, andalso to someone in France who shifted them into the now-standard GTHOR for-mat.

Anyway, I got to look through more games than just my own, and I thoughtI’d go through the transcripts, find an instructively badly played one, and an-notate it. But I didn’t find one: almost all the midgame moves in almost allthe games were sensible, thought-out moves. Not always good, but usually hardto criticise. But, an agonisingly large number of games featured a number ofendgame errors, ranging from the elementary to the bizarrely complex. Of course,many of these will have been due to time trouble blunders, and all I can say isthat, if you’re prone to these, then leave more time for the ending. (Easier saidthan done, I know.) So, to repeat, all of the positions below are the outcomes ofwell-played midgames, but disaster is about to strike.

© © © © © © ©• • • • © © © •• • © © © © • •• • © © © • • •• • © © • • © •• © • © • • • •• • • • • •• • • • • •Marson v. Haigh

First off, here is David Haigh, who had a goodtournament overall, but here he falls victim totwo careless moves. It’s worth getting to knowabout how to play out the three-square region inthe SE here. The rule of thumb is that you playwhichever of g7 and h7 does not permit h8: ifboth do, it’s often best to start with h8. Here,then White should be starting with h7. Should heplay 56b8 57g1 first? Give it a moment’s thought.Does it make any difference which order White’smove to b8 and Black’s to g1 come in? Yes, it

does: if Black moves first, then b8 turns the whole b-column. So White shouldplay 56h7. Which is Black’s better response? The answer is that he should playg7 first, to keep the d4 disc. Check this: it’s a help to be able to calculate thislittle detail quickly. Here it makes no great difference, but another time it couldcost you the game.

For the record, the unfortunate David Haigh, no doubt in the throes of timetrouble, played 56b8 57g1, and then compounded his error with 58h8 (P) 59h760g7 33-31 rather than 58h7 59g7 60h8 29-35. The difference is that, in the firstline, Black gets the last move, and the four discs between d4 and g7, whereasWhite gets these with the last move in the second line. White-Black-Whitealmost always works out better than White-White-Black.

16

• © © ©• © • © ©

• • © • © • © ©• © • • © © • ©© © © © © • © ©© © © © • • ©

© © • •• © © •Haigh v. Wormley

David Haigh again, to play at 45. Black, I wouldsay, is winning this easily enough. He’s obviouslyworried about the c6-f3 diagonal, and White’sthreat of g2. Should he be? Well, Black is hardlyin danger of running out of moves in any case, soyou might even say that g2 will merely lose parityfor White. E.g., 45g6 46g2 47d1 48b1 49b2 50h751g7, and Black would be winning even withoutthe swindle coming his way in the SW. I thinkit’s a good idea to play g6 before doing anythingelse, otherwise White is going to play there. You

should have in mind the sacrifice g6-h7-g7. Suppose the game goes 45g6 46h747g7 48a2. Now play 49d1 50b1 51b7 and the game will be over. It may not beeasy to calculate these lines, but Black’s advantages ought to be clear: in fact,all the regions on the board play well for him. So this is not the time to panic.Our beloved Secretary chose this moment to play 45b7. This is superficially veryattractive, as it seems that White will have to sacrifice to cut the diagonal. Fur-thermore, 46g6 47a7 is quite hopeless, so White has to play 46a2, met by 47g6,and now White is tremendously short of moves. This is in a way a nice tacticalshot, and it would be entirely in order to try something like this from a badposition, but here it is a big risk.

• © © ©© • © • © ©© • © • © • © ©© • • • • © • ©© • © • © • • ©© • • © • • • ©• • • • •• © © •

White to play

What should White do now? Again, this is notthe time for panic. What White did do was see away on to the diagonal, and take it. 48g8? Butafter 49h8, White discovered that being on thediagonal wasn’t all that much good, as taking a8would allow Black to wedge. Nevertheless, thisline (e.g., 50a8 51a7 52b2 53d1 54b1 55a1 etc.)is definitely still a better chance than 50d1, asplayed, which guarantees Black both h1 and g2at his leisure. Black won easily.

Back to the diagram. White’s best move isthe simple 48a7, challenging Black to find something to do. All there is is 49g2,and now White would really like to be on the diagonal, so 50g8!, with Whitewinning easily now.

17

• © © © •• © © • • ©• © • • © © © ©• © • • © • © ©• • • © • • • ©• • © • © • • ©• © • • © ©© © © © © ©

Wormley v. Selby

Matthew Selby (White) has just played to b2,and Mark Wormley (Black) has responded by tak-ing the a1 corner (this was in fact a mistake).Now, how is White to win this? This is quitea tough problem, requiring some accurate anal-ysis. Matthew played the routine sequence 52a253b1 54a8, and now Mark won with 55b7, buteven better would have been 55g7! 56g2 57b7 (P)58h1 59g1 60h8 42-22. So here is the question:how is White to stop Black gaining parity withg7 at the key moment? Playing 52g7 himself sac-

rifices too much, but here are two other thoughts: (1) organise matters so thatWhite plays one of the a2-b1 pair, and the response doesn’t reflip b2—then g7is not a diagonalisation; (2) try to play to b7 so that Black can’t reply with a8,pulling the same stunt that Black is planning in reverse.

Is 52b7 the way to accomplish these aims? No, since unfortunately Blackcan reply 53g2 (the “Principle of the Opposite X-square”), and White is nowdead. We are thus led to try 52g2. This works beautifully. If 53g7, then 54b7,so Black does better to play 53b7 himself, met by 54a8. Now 55g7 can be metby 56b1 (or indeed by 56g1), but Black can avoid this too by playing 55g1 56h157g7. Unfortunately for Black, these contortions to gain parity cost too much,and 58a2 59b1 (P) is 29-34.

• © © ©• © © © © © © ©• © • © • © © ©• © • • © © • ©• © • • © © • ©• • © © • • • ©• © © • • • • ©Brewer v. Haigh

Black has, reasonably enough, played to reachthis position. It looks an easy win, but let mestart by telling you that, without doing anythingoutrageous, Phil Brewer lost from here. I wouldstrongly recommend the following line: start with49b1 50a1 51c1 52d1 53c2 54d2 55e1 56f1 57e2 (P)58g1 59h1 60g2, then stop and count up. That is aclassic example of an “interior sweep”: Black pro-gressively gives White the top edge, meanwhiletaking all the discs in a huge triangle just belowit. Another advantage of the sequence is that it

costs no time on the clock. (Marginally ‘better’, in fact, is 55e2 56e1 57g2 etc.,but who cares? For the real Ebeneezers among you, the line from the diagramguaranteeing the most discs starts 49c1 50a1 51f1.)

Of course, there are other ways to win, but, especially when you’re Black, thegame won’t win itself. The general principle I’d like to state is that, no matterhow won the position looks, you should always have a plan, and always be tryingto achieve something.

18

• • • • © •• © © © © ©• • • © © • © ©• • • • • © • ©• • • • • © © ©• • • © • • © ©• • • • • •• © • • • • •Selby v. Brewer

Black (Matthew Selby) has gradually leaked awayhis advantage, and White (Phil Brewer, to move)is in a position to win. How? Phil played 54b2?,and now Black has a neat win with 55g7 56h857h7 58b1 59a1 60a2. In fact, Matthew played55a1?, after which White could have won with56h8 57h7 58a2 59b1 60g7. Instead, Phil played56b1?, after which Matthew had a win with 57g758h8 59h7 (P) 60a2. In fact, Matthew played57h7, which also won. But this isn’t OQ, so I’llgive you an explanation of all this. In a word:

parity! Whoever plays first into the SE region gets two of the three moves. True,White can get three of the four in the NW in the meantime, but only by unwedg-ing and giving Black the North edge (as well as the a1-h8 diagonal). So, Blackought to play g7 a.s.a.p., and White similarly with h8. No surprises, then, forspotting that 54h8 is the winning move.

After 54h8, 55g7 doesn’t work. White plays, for instance, 56h7 (P) 57b258a2 59b1 60a1 29-35. So Black plays 55h7, which not only takes the h-file, butalso puts a piece on the second rank ready to sweep across. This is now difficultagain.

• • • • © •• © © © © •• • • © © • © •• • • • • © • •• • • • • • © •• • • © • • • •• • • • • • •• © © © © © © ©White to play and win.

The usual way to treat a position with a one anda four square region, with the opponent havingaccess to neither, is to play into the four-squareregion, letting your opponent take the next twomoves in that region, meanwhile taking the singlesquare, and finishing with a great last move. Thistechnique, called ‘feeding’, is sort-of illustrated by56b1 57a1 58g7 59b2 60a2. But here the last moveis not at all so great, and Black enjoys a substan-tial victory. The other try for a feed is 56b2, butthen Black sweeps up with 57a2, taking the row

and destroying the feed. Alternatively, we can think of playing g7 first and thensorting out the four: something like 56g7 (P) 57a2 58a1 59b2 60b1. That se-quence makes sense, but reaps only 26 discs. In desperation, we are drawn toplaying 56a2! This looks unthinkable, as it gives Black access to the hole, but sowhat? If Black does play 57g7 (best, indeed), then we get three out of four inthe other region: 58b2 59b1 60a1 31-33. Meanwhile, 57a1 58g7 59b2 60b1 showsyou how a feed is supposed to work.

19

A Really Hard and REALLY Pointless Othello Puzzleby Aubrey de Grey.

There was a period of about six months last year, during which I had beatenImre Leader and David Shaman exactly once each in tournaments. The gameshad little in common for the most part, but they shared the feature that thekilling move which secured me the game was one that grabbed a main diagonalby moving to a C-square and flipping an X-square. I had (and indeed have)overlooked moves of this sort several times in other games, and I think it probablytakes many people some time to get used to looking seriously at such moves, butthey’re surprisingly often worthwhile.

So, rather than artificially introduce a genuine article about Othello withsome totally absurd and questionably relevant anecdote, I can now do the oppositeand artificially introduce an article utterly unrelated to good Othello play witha questionably relevant anecdote that might in fact be of some use to the reader.(If the previous paragraph led you to hope for some instruction as to when suchmoves are good, well, sorry.) The actual subject of this article is:

What’s the maximum number of times that a given X-squarecan be flipped in the course of a game of Othello?

You might like to try your hand at this yourself before playing throughthe following game, as many of the considerations that are involved will revealthemselves better by experiment than by a single example.

If you managed 14 or more, you did better than a certain current worldchampion who shall remain nameless; however, it is possible to get to 16. In thefollowing game, the moves with a star next to them are the ones that flip G7 (soyou’ll know quickly if you make a flipping error, which is unusually easy whenthere’s as much going on as there is here).

E6 F6 G6 G7 *G8 H6 F5 *H8 H7 G5*G4 *F8 *F7 *G3 D3 E8 *E7 *C3 H4 F4F2 D8 *D7 D6 F3 D2 H5 F1 E1 *G2*C7 C8 C6 B8 C5 B4 E2 C4 H2 *G1*B7 A8 D1 C1 C2 *B2 A2 H1 E3 B6A6 B5 A4 B3 A5 A3 *A7 H3 B1 *A1

OK, so 16 flips of the X square can be done. It’s quite easy to determine atheoretical maximum number of flips for any square, as follows:(1) For edge squares, all the discs involved in any previous flip (including thedisc placed and the one at the other end) must be flipped by the next flip. Thisgives a maximum of 0 for corners, 1 for C-squares, 2 for A-squares and 3 forB-squares.

26

(2) For other squares, the maximum is the number of squares at which placinga disc could theoretically flip the target square. This is the number of squares onlines through the target square, minus four (the initial piece on each line), minusany centre square on such a line that is not adjacent to the target square.

This gives the theoretical maximum values shown below:

0 1 2 3 3 2 1 01 17 18 17 17 18 17 12 18 20 20 20 20 18 23 17 20 23 23 20 17 33 17 20 23 23 20 17 32 18 20 20 20 20 18 21 17 18 17 17 18 17 10 1 2 3 3 2 1 0

Theoretical Upper Bounds

The fascinating fact is that, for any square otherthan the X-squares, one can very easily constructa game in which the square is flipped the numberof times shown above. The edges and the cen-tral squares are the easiest of all, but even thesquares labelled “18” should take you only a fewattempts. In fact, I’m told that a central squarewas flipped 23 times in a genuine game! If you trythis, you need only keep in mind that all movesto squares on lines through the target piece mustflip it, unless it’s the first disk on the line (not

counting the target square itself). This implies that the first move on any suchline must be to one of the two squares adjacent to the target square.

Which leaves the question: can an X-square be flipped 17 times? At firstI thought I might be able to prove that this was impossible, by analysis of themoves to neighbouring squares: for example, the 16-flip game above capitulateson the possibility of 17 at move ten, which doesn’t flip G7. But soon I foundways around such problems, and eventually achieved a way to get through mostof the game with the possibility of 17 still alive. Here’s the best I’ve been able tomanage, with moves that flip G7 marked by stars:

E6 F6 F5 F4 G7 *H8 G6 F7 *G8 H6E3 D3 C5 D2 F3 *C3 *G5 D6 C6 *F8*H7 *G4 H5 C4 *E7 *G3 C1 E8 *D7 C8C2 B4 H3 F2 B5 *G2 *C7 B1 H4 *B2H1 B6 H2 D8 A6 B3 *B7 *G1 E2 A4A5 A3 A2 F1 E1 B8 *A7

at which point we are defeated by asingle disk, F6, which if it were blackwould allow *A1.

© • • • • •• • • • • • • •• • • • • © • •• © • • • © © •• © • • • © • •• • © © © © © •• • • • • • • •© © © © © © ©

White to play.

Suitably infuriated, I have tried quite hard to find a way round this, but withoutsuccess. Accordingly I hereby offer the traditional year’s free membership toanyone who can find a 17-flip game or a proof that none exists.

27

Advertising Thor! by Joel Feinstein.

Do you have access to any kind of PC? Do you want to know what the World’s topplayers have been playing recently? Do you need a strong opponent to practiseagainst regularly? Do you need a tool to analyse your endgames for you and tellyou what you should have played? The French program, Thor, written by SylvainQuin, does all this and more.

As far as I know, Thor will run on any kind of PC, using any kind of graph-ics available (or text only, if necessary), and a mouse if there is one. The Thordatabase is regularly updated, and currently includes about 16,000 top gamesfrom the last twelve years. You can select from the database using many differentcriteria. For example, you might want to look at games played between DavidShaman and Imre Leader in 1992. Or you might want to look at all games follow-ing some particular variant of the Rose in the last five years, or some combinationof such criteria.

Once you have selected a collection of games, you can play back and forwardthrough them, and swap between them, or look at statistics showing you whichmoves were played most often, what percentage of games were won by each colour,and what percentage of games should have been won by each colour (accordingto perfect computer analysis with seventeen empty squares remaining).

If you want to play against Thor yourself, there are many levels of playavailable. Even its lowest level is a tough opponent, however. Thor will giveyou advice on which move to play if you wish. You can save any particularlyinteresting games, and see what would have happened if you had tried othermoves. You can also swap sides!

If you wish to analyse one of your games, you can enter the position youare interested in, or alternatively play moves for both sides to reach the desiredposition. Then give Thor some time to think, and let it analyse the endgamefor you. If you want some suggestions for moves in the midgame, a feature isavailable which will suggest which three moves are the most interesting.

Thor is an extremely useful tool for improving your play, and I recommendit highly. The program is available free of charge: just send a blank PC disc tome,Joel Feinstein, Mathematics Department, University of Nottingham, UniversityPark, Nottingham NG7 2RD,

or to David Haigh. The program comes with detailed documentation in English.

28

Bat out of Hell by Graham Brightwell.

Editor’s Note: The article “Could Frame Thy”, which many readers will havebeen expecting, is held over due to the existence of a more suitable title.

The latest in my series of articles on popular openings concerns the Bat. It’sabout time I did a Diagonal variation, and this is one that’s been causing mesome concern, due to the sudden popularity of a line that looks rather good forBlack. Having had a closer look, I’m now happy that (a) this line is good forBlack, and (b) there are plenty of alternatives. More of this later. Perhaps thereader should be warned that I’ve spent less time on this article than I’d haveliked, and I’ve made no attempt to consult other experts.

© •• © © © ©• • ©•

The Bat.

The position opposite arises after 1.d3 2.c33.c4 4.c5 5.d6 6.f4 7.b4. (Regular readers willwonder why I’ve departed from my usual orienta-tion. Good for them.) The opening has also beencalled the Cambridge, but that has now fallen intodisuse. It bears more than a passing similarity tothe Heath Bat, 5.b4 6.d2 7.d6, leading to the sameconfiguration of discs, but in a different positionon the board. The lines in the two openings are,naturally, similar for a few moves, but the assess-ments are different, and the reader is not encour-

aged to think that the right move in the Bat will work in the Heath Bat, or viceversa. Most strikingly, the most common move in the Bat is 8.b6, whereas theanalogous move (which also happens to be 8.b6) in the Heath Bat is not regardedas viable.

White has quite a lot of choice at move 8, with (in increasing order of popu-larity) b5, e3, e6, c6 and b6 all possible. I won’t be giving any strong recommen-dations, but the overall conclusion will be that several lines are fine for White.Of course, if you don’t intend to play the Bat yourself, all you need to know isone line which you’re happy with.

To start with, 8.b5 is not good. A typical continuation is 9.b3 10.d2 11.c612.b6 13.e2, with Black more than comfortable. Alternatively, White can try10.e2 11.c6 12.b6 13.e6, which is not obviously better.

A more interesting try is 8.e3. Now after the inevitable 9.b3, White canplay either 10.b5 or 10.c6. After 10.b5 11.e6 (11.d2 is also possible, but is lesstempting than after 10.c6), we reach the position on the left below. The positionon the right is from the Heath Bat, after the continuation 8.e3 9.e6 10.b5 11.b3.That line is generally regarded as good for Black, with the conclusion that 8.e3is wrong in the Heath Bat.

29

• © © ©• • © © ©© © • ©• •

The Bat.

©• © © ©• • © ©© © • ©• •

The Heath Bat.

Here, to remind you, is the Heath Bat analysis (from the diagram on theright above): if 12.c6, then 13.a6 is good, so White plays 12.a4, met by 13.a314.a2 (lines like 14.a5 15.a6 16.b6 17.c6 are regarded as good for Black, who hasa nice move to c2 in reserve) 15.a5 16.a6 17.f5, and Black should survive.

Now, what difference does it make moving the d2 disc to f4? It still seemsto be true that 12.c6 13.a6 is fine for Black. So 12.a4, and now 13.a3 14.a2 15.c2:this was not advisable previously because the d2 disc gave White access to a5.After 16.c6 17.a6, White’s position is already looking untenable. However, Whitecan again try 14.a5 15.a6 16.b6 17.c6 (this sequence, rather than some similarones, to keep a white disc at b4). Now, after 18.f7, White’s position is a lot moreviable than in the Heath Bat, because a move to c2 meets with a response to the(here, empty) d2 square.

If 10.c6, rather than 10.b5, 11.e6 12.b5 is far less appealing, so Black usuallyshakes the position up with 11.d2. After 12.e2, there have been quite a few gamescontinuing 13.f3 14.a3 15.a5. Now I’d recommend 16.c1 17.f2 18.c2, but Whiteusually finds something more imaginative, e.g., 16.c2. On balance, I think 13.b5,as in the Piau-Ralle illustrative game below, is better. I presume the idea was tomeet 14.b6 with 15.c7, leaving White with nothing clear-cut to do. Overall, thislooks fairly even to me.

©© © •• © • © ©• © ©• ©

8.e6 9.e3 10.c2.

Another not-so-popular line is 8.e6. This isanalogous to a Heath Bat line I recommended ayear or so ago, which immediately went out offashion. As in the Heath Bat, the theme here isthat 9.e3 looks like an ideal response, but thenWhite has the ungainly 10.c2 and none of Black’smoves seem to work very well (see opposite). Tobe brief, 11.b3 has to be met by 12.d2, whichdoesn’t deserve to work, but seems to. It oughtto be the case that Black organises access to b5and/or c6, plays there, and is ahead. But there

30

isn’t any simple way to get this access. 11.d2 can and probably should be met by12.d1, with similar themes. 11.f5 is perhaps best, intending something like 12.d713.c7 (gaining access to f6) 14.c6 15.b5 16.b6 17.f7. This is not all that decisive,but Black may have a slight edge.

Perhaps 9.e3 is not, after all the best, but the alternatives are relativelyuntried. Typical lines might be 9b3 10d2 11e3 12c2 or 9b5 10e3 11c2 12d7.

The two most common eighth moves are 8.c6 and 8.b6. Let’s start by com-paring the positions after (a) 8.c6 9.b5 10.b3 and (b) 8.b6 9.b5 10.c6.

© © •• © © © ©• © © ©© •

(a) c6-b5-b3.

© •• © © © ©• © © ©© © •

(b) b6-b5-c6.

There are two ways to think about the difference between these two. Youcould say that White has a disc either at b3 or b6; but you could just as well saythat White has a disc either at f4 or, after a reflection in the horizontal axis, atf5. Again, the two positions really are similar, and play can look just the samefor a few moves, but you are not advised to conclude that what’s good in one lineis good in the other.

For instance, the following line has proved devastating from position (b):11.f5 12.e3 13.d7 14.a4 15.a3 16.a6 17.a5 18.a2 19.b3 20.c2 21.f2. Experiencesuggests that Black is not now under that much pressure, that he will eventuallybe forced to flip the f4 disc, but that a subsequent white move to c7 can be metby a sacrifice against the West edge. The reader is invited to look for plausiblealternatives for White; you might try 18.b3 19.a7 20.e6, but I don’t think it works– Black can just play 21.f6, potter around in the South for a bit, then take the a2move with a lot of pressure (alternatively, he can go for an extraction, i.e, try toflip the b4 and b5 discs as well as play to a2, leaving almost no white discs buriedin the black mass). Another possibility is 12.f3 13.e3 14.e6 15.g4 16.d2 17.c2; thisdoesn’t look like anything to worry about, but it turns out to be astonishinglydifficult for White to find anything to do. To be frank, I don’t know what to doas White against this opening, and my advice (which I intend to follow) is to playa different move 8 and let someone else find a good way of coping with this.

So, is 11.f5 any good in position (a)? I suspect not. Part of the point

31

after 12.e3 in (b) is that Black is getting a move to b3 one day, but here thatsquare is occupied. Black should probably continue 13.b6, but White doesn’thave anything to worry about here.

Apart from the possibility of 11.f5, the main lines from (a) and (b) are justreflections: from (a), play goes 11.b6 12.e3 (White doesn’t want Black playingthere) 13.c2 14.a4 (14.a5 is just about possible, but not recommended). From(b), we get 11.b3 12.e6 13.c7 14.a5. See below.

•© • © ©

© © © © © ©• • © ©• • •

Line (a).

• • •• • © © ©© © © © ©© • © ©•

Line (b).

•© • © ©

© © © © ©• • © © ©• • •

Line (b), reflected.

As we’ve said, to choose between these positions, look at, say, the first one,and decide whether you want a white disc at f4 or f5. To be definite, let’s workwith (a). Black has two real options (15.e2 16.d2 appears to be fine for White).

• • • ©• • © ©

© • © • © ©© © © • •© © • • •©

After 21f6.

First, there is 15.a5 16.a6 (it’s quiet, yes, butso what?) 17.d2 18.a7 19.a2 20.e2 (note White’smove order: e2-a2-a7 instead leaves Black an easymove to f1) 21.f6, reaching the position opposite.Now White is currently short of good moves; onepossibility is 22.c1, leaving Black the long termproblem that he won’t be able to sacrifice withb2, since he’s unlikely to have access to a3 afterWhite takes a1. It’s probably close.

Now shunt the f4 disc to f5: much better forWhite, since he now has 22.c7. This is then a plus

for 8.b6 over 8.c6.* * * * * * * * * *

Ian Turner, along with almost everyone else, has so far failed to come up withan Othello Theorem. However, he has come up with a Hypothesis: “Reversi wasdeveloped from an earlier gambling game involving turning over coins trapped bythe one played. The thoughts behind this are: (1) Coins are two-sided, similar toOthello pieces, (2) Reversi has the strange rule not allowing pieces to be passedto the opponent when he has run out.”

32

• •© • • ©

© © © • © ©© • • ©

© • • •

Line (a): 15d2 16a6.

Alternatively, from the leftmost of the threediagrams above, Black can play 15.d2 immedi-ately. Comparison with the previous line suggeststhat 16.e2 17.a5 is going to be good for Black, butWhite has the interesting shot of 16.a6: see oppo-site. The simple idea is that White is not goingto lose a tempo on this edge: indeed Black shoulddelay the move to a5, to make it as awkward aspossible for White to subsequently get access toa3. So Black should chip away at the White for-mation to the East. How well this works depends

on what that formation actually is. So, do you want a White disc at f4 or at f5?Before 11.f5 came along, the consensus was that White wanted the disc at

f5, so that 8.b6 was a better move than 8.c6. But right now I can’t reconstructthis reasoning. From the diagram, play usually goes 17.f2 18.f3 19.a5 20.d7 (loud,but clearly worthwhile), and now Black chooses between c7 and g3. It’s close.

Shunt the f4 disc to f5 again. What’s so bad about playing to f2 here? Inpractice, Black almost always goes to f6, unpoisoning the c7 move for White. Onthe other hand, maybe White would prefer the c7 move to turn the disc at d6,to ensure access to a3 after a Black response to a5? Anyway, the sequence fromline (b) is long and well-travelled: the Tamenori-Rose illustrative game is verystandard, and there are at least two other major tournament games following thesame line to move 38. I believe Black is winning at that point, but I don’t believethat proves anything. The standard belief is that this line is good for White.

Conclusions? Well, I won’t play 8.b6 for a while because I’m scared of the11.f5 line, but I don’t see why I shouldn’t play 8.c6. (Or 8.e6, or 8.e3.) If you aregoing to play 8.c6, you also need to be aware of 9.b3: see the Marconi-Murakamigame for a typical continuation.

34 17 14 16 15 19 25 3835 31 18 11 12 20 47 4622 9 2 1 8 26 27 4823 7 3 © • 6 32 4928 13 4 • © 43 37 3329 21 10 5 40 39 44 5030 36 52 24 42 41 58 5945 55 56 51 53 54 57 60

Piau - RalleWorlds 1990, Semi-Final

47 48 32 26 31 30 60 5944 41 10 25 29 33 42 4938 27 2 1 9 21 20 5039 7 3 © • 6 24 3640 15 4 • © 11 19 4322 16 14 5 8 28 37 4553 52 13 12 18 17 56 4654 55 34 35 58 23 57 51Brightwell - Feldborg

Karsten Unrated 1988

51 52 24 37 50 48 53 5918 42 20 23 25 21 60 3915 19 2 1 12 22 32 3614 7 3 © • 6 33 3817 9 4 • © 11 31 3416 8 10 5 30 26 27 3547 41 29 13 28 44 56 5846 54 43 49 40 45 55 57Svirskiy-SenchevLeningrad 1991

33

56 57 58 59 60 27 52 4719 49 13 17 22 35 42 5344 10 2 1 12 26 36 3414 7 3 © • 6 33 3815 9 4 • © 23 37 5416 11 8 5 21 31 28 5118 40 43 25 20 24 39 5545 46 30 41 32 29 50 48Marconi-Jensen

Worlds 1990

47 41 45 46 40 39 60 5948 42 20 22 24 27 50 3716 11 2 1 23 17 29 3421 7 3 © • 6 19 3214 9 4 • © 30 26 3328 8 10 5 12 38 36 3153 49 13 15 18 35 44 4352 54 55 56 57 25 58 51Tamenori-Rose

Worlds 1990

43 41 23 32 28 33 60 5944 42 21 10 20 51 48 5522 9 2 1 17 18 47 4614 7 3 © • 6 19 5031 12 4 • © 15 36 3740 27 8 5 11 29 45 3858 49 13 16 26 24 56 3954 53 30 25 34 35 57 52Murakami-Marconi

Paris Open 1991

The Art of Sacrificing Four Corners by Joel Feinstein.

This article is intended to be a beginners guide to some strong ways to sacrificecorners against players who take a lot of pieces and a lot of edges (certain poorOthello programs, for example). But many of the ideas can be used in ordinarypositions also.

The first example we shall consider is the one I usually use to demonstratethat having lots of pieces and lots of edges is not a good thing.

© © © ©© © © ©

© © © © © © © ©© © © • • © © ©© © © • • © © ©© © © © © © © ©

© © © ©© © © ©

Black to play

In this fairly extreme example, White has a posi-tion which may seem ideal to many beginners. Hehas 44 pieces to Black’s 4, he has the beginner’sfavourite edge structure on each edge (a balancedfour), and it even looks fairly certain that Blackwill lose all four corners. But, this position is avery easy win for Black. Black has a particularlystrong type of sacrifice available. Black’s best linebegins with an x-square, e.g. 45.b7. White nowhas the choice of taking the corner, or giving itto Black (by playing a7 or b8). The latter choice

is, at least in this case, clearly suicidal, as Black can then proceed to win allthe corners and edges. So let us assume that White takes the corner with 46.a8.Black can now play both the moves a7 and b8, and should do so. This givesBlack wedges in two edges, more than enough compensation for his corner. Thefeature that made this sacrifice so good was that all of the pieces in the b columnand the 7th row were white. Note that this sacrifice can be used in the midgamein order to play three out of four moves near a corner, and force White to findtwo moves elsewhere. In this endgame, however, White simply passes, and thereis more work to do.

After 47.a7 pass 48.b8 pass, Black plays 49.g7. This time the pieces on the

34

seventh row are not White, but after 50.h8, Black has access to g8 along thebottom edge, and so we have 51.g8 pass and black can save h7 for later. After52.b2 a1 a2 pass Black has two excellent moves saved up: he only needs one tomake the last region play the way he wants. In order to get the last move in theNorth-East region, Black “feeds” White with some self-destructive moves. 55.g1h1 h7 g2 b1 pass h2 completes the perfect play line. Black has sacrificed all fourcorners and won 46-18.

© © © ©© • © ©

© © © © © © © ©© • © • • © © ©© © © • • © © ©© © © © © © © ©

© © © ©© © © ©

Black to move.

Let us see another example. This time three ofthe x-squares available to Black are inferior tothe fourth. Since the pieces in the g column andthe seventh row are all White, the best first movefor Black is 45.g7. Now, as before, perfect playcontinues with 46.h8 g8 pass, but this time Blackwill need to save his available move at h7, and feedWhite immediately in one of the other regions.Best is 48.b1 a1 h7 b2. Note this has made allthe second row White, and so g2 now becomesavailable as a good sacrifice. After 52.g2, White

should consider h2, but 53.h1 h2 pass g1 pass is better for White. Note Blackstill has his move at a2 saved up, and so he can feed White in the last region.Surprizingly, the best way is to unwedge with 56.b8, giving White the fourthcorner, and the whole South edge. But after 57.a8 a2 b7 a7 Black wins 43-21.

When there are no obvious strong x-square sacrifices, the wins in this type ofposition are hard to find, if they are even there. I would suggest that you shouldtry to avoid ending up in the following kind of position, either by making surethe opponent’s edges are worse than this, or by taking an edge yourself at somepoint.

© © © ©© • © ©

© © © © © © © ©© • © • • © • ©© © © • • © © ©© © © © © © © ©

© • © ©© © © ©

Black to move.

I will admit that when I typed this position intomy computer, I thought it was a win for White!But it turns out that Black can win 33-31. Theperfect play line begins 45.b1 a1 a2 b2 b7 b8. HereWhite turns down the corner in order to cut ac-cross Black’s plans. But Black simply continueswith 51.a8 a7 g8 g7 h8 h7 h2 g2 g1 h1, and wins,barely, 33-31. What happens if White grabs cor-ners instead? After 50.a8, Black sacrifices with51.h7 and perfect play is then 52.b8 g8 h8 g7 g1(feeding time again) h1 a7 g2 h2 34-30. If instead

52.h8, then 53.g8 g7 b8 pass g1 h1 a7 g2 h2 is 35-29 to Black.I would not like to have to find that over the board.

35

Now, for a change, let us look at a real position from a serious game.

© © © ©• © • © ©• • © © © © © ©• © • • • © © ©© © • © © © ©

© © © © © © © ©• © © ©• • • ©

Bhagat (Black) to move.

This position is from Bhagat-Feinstein in the 1987Cambridge international tournament. I wouldn’treally want this position for either side! Pete hadoutplayed me in the mid-game, and in despera-tion I had walled myself off from both regions inthe East, after making sure that he had no goodx-square sacrifices. We had now arrived in theendgame with Black to play. Pete tried 45.b2,(this is indeed a 33-31 win). But after 46.a1 Peteplayed the natural 47.b1 and after 48.a5 I waswinning all the way. Black urgently needs a plan

from the start, and it is very difficult to find one.The perfect play line is 45.b1 a1 a5 b2 g1 g2 b7 a8 a7 e8 h1 h2 h7 g7 g8 h8

37-27, but the question is: what happens if White plays 50.h1 instead? Well, it’svery sneaky. Black plays 51.b7 a8 e8 a7 and Black gets both of g2 and h2. Playfinishes 55.g2 pass h7 (feed) h8 h2 g7 g8 and Black wins 38-26. If you saw all ofthat, you are an endgame expert!

© © © ©© • © ©

© © © © © © © ©© © © • • © © ©© © © • • © © ©© © © © © © © ©

© © • ©© © © ©

Black to move.

Finally, a tricky little exercise for you. Would yoube able to find a win for Black in the followingposition? Black can win 33-31.

(Warning: after 45.b1 a1 a2 b2 b7 b8 Whitegets a draw. So you need another plan).

A solution (well, clearly there is more thanone!) can be found on p.44.

Rome International Tournament.Marc Tastet reports that he has lost another Grand Prix Final, his third thisyear. [Editor’s Note: presumably this puts him out of contention to defend hisWorld title?] Marc lost 2-0 to Francesco Marconi, while Stephane Nicolet beatDominique Penloup 2-0 for third place. The final result, claims Marc, is verysimilar to that in Cambridge! Donato Barnaba, Enrico Colanguiolo and AndreaSilvola were tied for fifth.

Last year, it has been pointed out that Graham Brightwell established a hard-to-beat record of won games within the same year in the finals of European GrandPrix tournaments (6.5). Marc suggests that he (Marc) has already established arecord of lost games with 6, and 2 tournaments still to come!

36

The 1993 Cambridge Open by Karsten Switness.

A good turnout of 26 for the latest in this series of prestigious tournaments. Itwas particularly good to see three Polish players: Pawel Pietruszkiewicz has beentheir top player for ever, it seems, but he was a little out of form in this event.The French sent, as usual, a strong contingent, headed by World Champion MarcTastet. Erik Jensen played the key role of tournament Dane, and distinguishedhimself by falling for an 18-move wipeout against Bintsa Andriani.

The first round saw no British player lose to a foreign player, with Ian Turnerand Phil Marson scoring particularly good wins over Stephane Nicolet and Do-minique Penloup (two of the top French players) respectively. The French cameback strongly however.

Imre Leader led overnight with 6/7, half a point ahead of David Shaman,but slumped badly on the second morning. Going into the last round, Shamanwas a half point in front of Tastet and Nicolet. Tastet is well known for alwayshaving a good tiebreak, so it looked like a Shaman v. Tastet final. But thenTastet lost to Emmanuel Caspard, leaving the door open for others. Nicoletcould have made it with a win over Brightwell, but he lost. Leader presumablyhad a decent tiebreaker, and might just have made it with a win over Bhagat,but he lost too. So, Shaman v. Tastet it was, a repeat of last year’s World Final.The difference is that British players can’t win World Finals (Shaman, indeed,didn’t), whereas French players can’t win Grand Prix tournaments (Tastet, in-deed, didn’t). [Editor’s Note: This explains Shaman’s recent decision to becomeFrench, one supposes.]

Name Pts GP Name Pts1. David Shaman 9 1

2 +2 200 13. Serge Alard (B) 52. Marc Tastet (F) 8 +0 140 Joel Feinstein 53. Stephane Nicolet (F) 8 +2 90 Erik Jensen (DK) 54. Emmanuel Caspard (F) 7 1

2 +0 60 Dominique Penloup (F) 55. Graham Brightwell 7 35 Piotr Pietruszkiewicz (PL) 5

Imre Leader 7 35 Guy Plowman 57. Peter Bhagat 6 1

2 20 Matthew Selby 58. Bintsa Andriani (MAD) 6 6 Ian Turner 5

Iain Barrass 6 6 21. Aubrey de Grey 4 12

Alexandre Cordy (F) 6 6 22. Roy Arnold 4Mike Handel 6 6 Phil Marson 4Pawel Pietruszkiewicz (PL) 6 6 Witold Postrach (PL) 4

25. Colin Hands 226. Simon Nickson 1

37

Two Bees with One Stone by Marc Tastet.

Here are some comments on two games I played in a row, rounds 6 and 7, on thefirst day of the ’92 World Championships in Barcelona. These games illustratehow two of the so called “killer bees” where killed themselves using the sameopening, which might explain the title. [The Handel-Tastet game doesn’t fitin with this picture of reality, and hence is not mentioned. The author seemssomehow to have got the impression that this is normal practice. – Ed.]

Let’s start with the Tastet-Feinstein game from round 6.

51 52 58 59 34 20 49 4845 46 36 35 33 19 37 1728 22 32 13 3 6 12 1630 27 7 © • 2 10 1543 18 5 • © 9 11 3929 38 14 4 1 8 21 4447 41 31 26 25 23 42 6050 57 40 56 55 24 54 53

Tastet 44-20 Feinstein

1e6 to 10g4: The 8f6 move is the Rose opening, avery classical one after which moves 9 and 10 arealmost always played this way.11g5: This move is the Greenberg variation, afterthe name of an American player who used it in1983 to defeat Brian Rose. This variation was abit out of fashion because it was considered as alosing opening for Black. It made a spectacularcome-back at the ’92 World Championships be-cause both the Italian player Francesco Marconiand I independently decided to play it again.

12g3 to 16h3: Very classical continuation after the Greenberg. You can reach thesame position after move 15 by switching moves 12 and 14.

• • © © ©• • • • © •• © © © •© © © ©

After 16h3.

17h2: Here is the novelty, the standard movebeing 17g6. I learned this move when VladimirPolyakov (whose name I had never heard before)played it against me and beat me in last summer’sChelyabinsk Open. (By the way, if you have theopportunity to travel to Russia one year, don’tmiss it! You can get details from Imre Leader,who was there with me. We both very muchenjoyed our stay there.) In Barcelona, Marconiplayed 17d7 which might also be an interestingvariation. Maybe the Greenberg had been buried

too quickly?18b5: This is one of the two standard replies, the other one being h5.19f2 to 21g6: Playing the pair f2 f1 before playing g6 has the advantage ofinstalling a white disc at f4 so that if White answers 22h6, he also flips g5 andloses access to h5, which wouldn’t have happened had Black played 17g6 or 19g6.22d7 and 23f7: Very natural moves. We had reached the same position in thePolyakov-Tastet game from Chelyabinsk and we’ll reach it again in the next game.

38

©© •

© • • © • •© • • © © •

© © © © • •© © © • ••

After 23f7.

24f8 and 25e7: White threatens to play h7 fol-lowed by h5 and Black prevents it.26d7: In Chelyabinsk, I had played 26e2 but I wascompletely dead out of the opening; Joel’s moveis probably better.27b4 to 31c7: Black needs to win a tempo some-where. I was very happy to find the elegant se-quence from 27 to 31 achieving this aim. The keymove of this sequence is 31c7 which gives Whiteaccess to b6 but simultaneously poisons this movebecause now it would also flip c5, d4 and e3.

32c3 to 36c2: Move 32 is a bit surprising, but is consistent with Imre Leader’sflat wall theory. Anyway, it’s not easy to find moves for White.

© ©© © © © •

© © © © • © • •© © • • © • © •© • © © © •• • • © • •• © © ©

©After 36c2.

37g2: This move didn’t surprise Joel. Of course,it would be stupid to try to surprise Joel playingan X-square! If you want him surprised, don’tplay an X-square because this is the first move heexpects!38b6 and 39h5: One of the aims of 37g2 was toget access to h5. So, maybe at 38 White shouldhave played h5 himself!40c8 to 44h6: At first sight, White’s play doesn’tlook very consistent, because first he doesn’t wantto take the h1 corner at 40 and then he sets a

Stoner trap to get it! However, there is a big difference: if White takes the cornerat 40, he loses access to g1 and might lose parity there because Black can waitto play in the hole. After 41b7, Black controls the diagonal and White won’t flipg2 when he plays h1, forcing Black to reply with g1.45a2: This move is not risky because b6 is white, so that when White plays a8,he won’t have access to a7.46b2: White has only two sensible moves, h1 and b2 but he chose the wrong one;46h1 loses only 39-25.47a7 to 50a8: Black chooses a sequence which semi-forces white moves (andhappens to be the best one). Black is in full control of the game.51a1: This move did surprise Joel because he missed the fact that my move 49had flipped the disc at d4 and he thought I had no access to the corner. Thoughhe had only one legal move at 52, he spent about half a minute before playingit because he was trying to understand how he missed that and then convincedhimself that 46 was a mistake.

39

• © © © • ©• © © © © • © ©• © • © • © • ©• © • • © © • ©• • • © © © © ©• • © © © © © ©• © © © © © ©© © ©After 52b1.

53h8 and 54g8: Both moves are non-optimal, (atleast if the objective is to have the greatest num-ber of discs at the end). In this sense, 53c1 and54h7 were better. 53h8 is wrong because it al-lows White to keep discs on the b1-h7 diagonaland 54g8 is wrong because it gives back the samediscs to Black!55e8 to 60h7: Perfect endgame. White has nochoice and Black plays so as to have all the c col-umn black in case it’s better to play 58d1 insteadof c1 but it isn’t.

Now it’s round 7 and I have to play David Shaman, who has won all hisgames so far. I’m second alone with only one defeat (vs Sakaguchi) and severalplayers have 4 points out of 6.

58 59 60 52 51 20 45 3856 57 44 46 53 19 35 1730 22 36 13 3 6 12 1632 29 7 © • 2 10 1547 18 5 • © 9 11 3931 49 14 4 1 8 21 3448 50 33 28 25 23 41 4354 55 40 27 37 26 24 42

Tastet 41-23 Shaman

I’m Black again, and as it worked well with Joel,I play the same opening and the game plays thesame way to move 23, so you can start from thediagram above after move 23.24g8: This is a trendy move (see Glossary in NewYork, New York). It looks bad at first glancebecause White might lose the h8 corner after thismove. But actually, it isn’t that surprising: Blackhas potentially a five on the eastern edge becauseWhite can play h5 when he wants and Black mustreply h6 or lose the h1 corner. If Black threatens

to take h8, White can give Black the five and then he’ll be able to wedge at h7if Black takes the corner anyway. However, it turns out that 24g8 is a bad moveafter all.25e7 to 28d7: So, I don’t try to take the h8 corner, but I leave White without amove on the southern edge and with a weak position there.29b4 to 33c7: At this point, I realise that the position is very similar to what itwas in the previous game at move 26. Actually, the only difference is that thereare two extra discs: a black one at d8 and a white one at g8. But they don’treally change the position and I can solve my problem (win a tempo) in exactlythe same way. So I play exactly the same sequence.

40

©© •

© © • • © • •© © • • • © © •© • © • © •• • • © • •• © © ©• © ©

After 33c7.

34h6: David doesn’t want to open the northernpart of the board. 34b8 would be met by 35b6,so White finally plays 34h6, probably a bit reluc-tantly.35g2: This is a thematic move, given the positionon the eastern edge: if 36h1, 37h5 and Black haswon a tempo (and maybe also parity with g1 be-ing an odd hole where White cannot play). It’svery important that White has no access to h5,otherwise he could play there before taking theh1 corner and Black would have lost everything:

he couldn’t wedge on the eastern edge. It’s a classic mistake that might cost youa game some day (if it hasn’t done already), if you don’t take care.36c3: This move takes advantage of the fact that 35 flipped f3 which was poisoningc3. However, it leaves Black an easy reply.37e8: I’m preparing a nasty surprise for White at move 40. I must admit Ihadn’t seen it coming when I played 35g2 and certainly David hadn’t either,since otherwise he wouldn’t have played 36c3. You can try and guess what thesurprise is.38h1: This move is nearly forced if White doesn’t want to play c8 and close allthe south-west region. On any move other than these two, Black can just takeh8.39h5: It’s much better to wedge than to take h8 corner. Can you see the surprisenow?

© ©© © •

© © © • • © • •© © • © © © © •© • © • • • •• • • • © • ©• • • •• • © ©

After 39h5.

40c8: White probably wanted to play 40g7, win-ning a much needed tempo and keeping parity inthe south-east region, but 41h8 would leave himwithout access to h7! Moreover, 40h7 41h8 leaveshim without access to g7! As he cannot play inthe south-east region, White is forced to play c8,otherwise I take the h8 corner and it’s even worse.Finally at move 40 we discover that move 24 wasbad, but of course it was impossible to foresee.Let’s say that 24g8 was a Shaman move (see againGlossary in New York, New York).

41g7: makes the win safer by winning a tempo in the south-east region andpreventing White from doing it himself.42h8 and 43h7: I didn’t think White would be in a hurry to play this pair outand give me definitely the eastern edge. He could have waited, hoping that paritywould force me to play h7 and give him the edge, though this is not very likely as

41

White has lost control of the game. However, in the best sequence, found with acomputer, White does play the last move in the south-east region, but it’s Blackh8 followed by White h7. So, 42c2 loses 41-23 after g1 e2 a5 a7 b6 b7 a8 b8 h8h7 d1 e1 a2 c1 d2 b2 a1 b1, White keeping parity everywhere.44c2: Semi-forced: it’s the only move that flips only one of the two discs in theNorth.46d2: e2 saved two more discs (43-21) after playing a sequence very similar towhat happened in the game.48a7: b6 kept parity for White and saved two more discs (45-19) after 49a2 b2a1 b1 c1 d1 b7 a7 e2 e1 a8 b8.

© • ©© © • • •

© © © © © © • •© © © © © © © •© © • © • • • •© © © • © © • •© © © © • • • •

© © © © © ©After 50b7.

51e1: This choice might look surprising at firstglance: why does Black, being wedged with g1,choose to lose his wedge? It’s to win parity! After52d1 and 53e2, White passes and there is an oddhole in the north-west where Black will play thelast move. After 51e1, the endgame is perfect:White has no more choice and Black wins 41-23.Actually this move 51 is not that good for tworeasons. The first one is that Black can have asmany discs without parity playing 51e2 e1 a8 b8a2 b2 57a1 b1 c1 d1 also winning 41-23! Notice,

however, that as Black didn’t have parity he couldn’t use his wedge at g1 to getdiscs on the northern edge. The second reason is that Black can improve theprevious sequence so as to win parity: instead of 57a1, he can take advantageof the b column which is all white and play 57b1! White answers with one ofc1 or d1, it doesn’t matter, Black takes the a1 corner, White passes and Blackplays the last move of the game, winning 47-17. This way, Black got the wholenorthern edge (except of course h1) thanks to parity and his wedge at g1. Thisis very classical: such a wedge is useful only when you have parity.

This was to be David’s only loss in the 13 rounds of the Swiss system.

* * * * * * * * * *

In the European Grand Prix, Tastet leads with 420 (140+140+140), followedby Shaman 290 (200+90) and Feinstein and Marconi 200. Nicolet has 180 andPenloup 100. Only the best three scores count, so in some sense Shaman’s positionis better. If he wins either Brussels or Paris, then he wins the Grand Prix; if hemanages only a second place, then Tastet needs to win one of the two events totake the coveted title. A win in Brussels for Feinstein, or Marconi, or Nicolet,would make things even more exciting.

42

Syncopated Cerebrations by Sid Cox.

At last, a theorem. Alas, neither I nor the author can see any relevance in it toactual Othello play. So I have decided to wait until the end of the year, and if ithas no rivals by then, the prize is yours, Phil.

What about this then. Do you think it might be true? Here, with apologiesto Arthur C. Clarke, is Sid’s First Conjecture:

“Any sufficiently advanced Reversi game is indistinguishable from Othello”.By this I mean that, starting from the non-Othello initial position of the four

discs in Reversi, a situation will always eventually be reached which can also bereached from the true Othello starting position. Ignore the Reversi rule aboutunplayed disc ownership if you like.

Actually, those of you who read the OQ and the Scientific American mayknow a counter-example which disproves this conjecture, so an additional con-dition must be imposed. This is: “Games ending with empty squares do notcount”.

How soon can you get a Reversi game and an Othello game to converge?

+ + +

Meanwhile, Aubrey the Achiever has produced a number of curious facts aboutOthello. Apart from his work on the maximum possible number of moves, shortestnon-wipeout games, shortest draws and the maximum number of times a piececan be flipped (see his own article), he told me that he had played a game withColin Hands in which one of the players made eleven successive passes. Can anyof you beat this record? This leads nicely on to the next topic.

+ + +

From time to time questions are asked about handicapping Othello games, some-thing which does not happen in BOF tournaments, but which might be useful forprivate games. Several ways of doing this have been proposed.

Handicapping by giving 1, 2, 3 or 4 corners to the weaker opponent is sug-gested by the instructions provided with some Othello sets, but has the disadvan-tage of changing somewhat the character of the game. The corners are just toocrucial to be used as handicaps, except perhaps between players of very differentabilities.

Handicapping by awarding disc points to the weaker player has also beensuggested, and would allow finer control of the handicap, but suffers from thedisadvantage that disc differential is not a very good indicator of the relativestrengths of the players. The effect of this sort of handicap would not be consis-tent from one game to another; it would be difficult to know what handicap touse between players of very different ability. Also, like the corner handicappingmethod, it doesn’t teach the weaker player anything.

43

The best handicap I have heard of is to give the weaker player the optionof passing once, or twice . . . or N times during a game. This does not alter thecharacter of the game, and has the advantage of emphasising to the weaker playerthe importance of not running out of good moves. (Think of the number of timesyou have said to yourself when in a bad position “if only I didn’t have to movenow”!) Making the strong opponent move twice is a better compensation for thebad position the weaker player has got him/herself into, because it usually resultsin the position being improved for him/her. Also, learning when to use the freepass teaches the weaker player to recognise crucial stages in the game.

At this point I cannot resist claiming the kudos for being the first to publishthe shortest possible handicapped game. It is:

F5, pass; C4.

Right now I do not know what advantage being allowed a free pass confersupon a player. Is one free pass equivalent to 200 rating points? It must dependon the ability of the player; in the hands of an expert one free pass could bedevastating, even against a grand master; in the hands of a novice several freepasses (inappropriately used) may have little effect, or even (as above) the wrongeffect. So the best advice I can give is to start with one free pass, play a fewgames, and if the weaker player still loses consistently increase the number of freepasses. You just regulate the number of free passes so as to keep your gamesrelatively even as the weaker player improves.

It would be nice if writers of Othello programs could include the abilityto do handicapping like this. People would learn to play better Othello if thehandicapping was done in this way, rather than (or in addition to) the presentmethod of making the computers play more stupidly.

Can any of you think of any snags with this handicapping method? Is thereany problem with allowing the handicapped player to use two or more free passesin succession if he/she wants to?

* * * * * * * * * *

Othello Programs.You can still get three good Othello programs for the PC for free by sending mea formatted diskette (either size) and a SAE to David Haigh.

* * * * * * * * * *

Solution to puzzle on p.36. Here is an optimal sequence: g8 h8 b7 a8 g1h1 h2 b8 (if g2 then h7!. White is fortunate that Black has no access to g2) a7(Swindle coming!) g2 h7 pass a2 (big feeding sacrifice) a1 g7 b2 b1 (just!) 33-31.

44

Correspondence edited by Graham.

I thought my article on Qualification Systems would raise some hackles, butnary a murmur. So either everybody accepts my descriptions as accurate, or I’mbeneath comtempt.

I did get a number of favourable comments about the endgame challengerunning through the last newsletter. These will happen again, sometime whenthe deadline is not so tight. Unfortunately, I also received the following letterfrom Wayne Clarke of Hartlepool (lightly edited).

I enjoyed your “endgame” puzzle – this would make an excellent regularfeature. However, I want to question your comment on page 35 that White’s52c2 followed by Black’s 53g7 “loses horribly”. It seems to me that 54b4 is still awinning line for White, and produces a bigger win than your preferred line whichleads to a 34-30 win.

Black’s best response to 54b4 is 55h8, which is followed by 56c4 57b2 58a259a1 60b1: White wins 35-29. If Black plays 57b1 then 58a1 59b2 60a2: Whitewins 37-27. If Black doesn’t play the pair to the h8 corner and plays 55c4, thenplay goes 56h8 57b2 58b1 59a1 60b1: White wins 39-25.

Something else I discovered is that if 52c2 is countered with 53h8, thenalthough this seems not so good for Black as 53g7, it works out to the sameresult – the best line seems to be 54g7 55c4 56b4 57b2 58a2 59a1 60b1: Whitewins 35-29.

I know I am a mere “beginner” but 52c2 does seem to work for White, andI think you dismiss it too quickly.

By the way, does anyone know how the Othello Game Boy Cartridge plays?I’d love to get an opinion I trust before considering buying it. I had a pleasantsurprise with the Reversi program that comes free with “Windows” – it’s quitestrong and is worth taking on.

I believe that others have a lower opinion of the quality of Windows Reversi.Apart from that, I can’t raise even a quibble. It still seems strange to me thattaking three moves in a row into the big region can possibly be enough compen-sation for giving up both h8 and g7, but definitely it is. I must have been soconvinced that this was hopeless that I didn’t even bother letting a program lookat the initial position, which is, erm, careless. Interestingly, nobody else seems tohave spotted this. Do you all trust me? (My track record is not good!) Oh well.Noises off. Brick wall colliding with Editor’s head. Cries of anguish. Cacklingex-Brightwell fans moving in for the kill.

45

To Rate or Not to Rate? by David Haigh.

Which tournaments are rated? The committee has recently decided that if atournament is going to be rated by the BOF it must fulfil the following fourconditions:(1) It is played in the UK.(2) British players are eligible to play in it or to qualify to play in it.(3) It is played with clocks, with at least 20 minutes per game for each player.(4) The players know that the tournament will be rated. This is assumed to have

been the case for Regional, Final and International tournaments in the past,and notice is hereby given that in future these tournaments will be rated.Conditions 1 and 2 ensure that the tournament is within the jurisdiction

of the BOF. Conditions 3 and 4 ensure that the tournament is taken reasonablyseriously. “Friendly” tournaments such as those held in Cambridge and Doncasterare sometimes rated and sometimes not, according to the wishes of the playersand the organiser. Interestingly, the 1993 World Championships will satisfy atleast (1)–(3).

Some tournament games are not rated. These include bye games and gameswon by default, for example where the opponent does not turn up. Very occa-sionally there is a game whose outcome is in doubt, and when I am aware of thisdoubt I have not rated such games either.

The first such occasion was in the 1988 Cambridge International, in a gamebetween Joel Feinstein and Marc Tastet where Joel managed to destroy the boardposition in his haste to flip the discs. It was thought that Joel had won on theboard, but there must have been a flipping error as the remembered positioncould not be re-created. Joel probably should have lost as a penalty for messingup the position, or because he would have lost on time if he had been made torestore the board position. Marc generously offered a draw. In the face of suchconfusion the outcome of the game seems indeterminate and I judged it best topretend that the game had not taken place. The poor referee does not have theluxury of ignoring a game and in this case accepted Marc’s offer and declared thegame a draw.

Now it is extremely unlikely that the real outcome would have been a draw,and to have used a draw in the ratings would almost certainly have distorted them.In my view Joel should have lost, and his rating should reflect his difficulty incontrolling his discs, but to have rated the game in this way would have beenarrogant and irresponsible. I therefore decided that the best course was to ignorethis game.

The second occasion was a game in another Cambridge International, be-tween Peter Bhagat and Phil Marson this year. Here, both players ran out oftime and no-one knew (or admitted to knowing!) whose flag had fallen first. For

46

the lack of any laid-down procedure for handling this situation this game was also(rightly, in my opinion) declared a draw, which again it almost certainly wasn’t.This was an even clearer case of genuinely not knowing the real outcome of agame, and I decided not to rate this game either.

Fortunately for me, the third and most recent example shows that I amcompletely impartial in making these judgements. The scene is again Cambridge,in this year’s regional when I played Iain Barrass. It had been a good, excitinggame (i.e., I had survived into the endgame) but I had not done well enough toprevent Iain having parity, and he got the last move in the last three regions. “Ahwell”, I thought, “I nearly won that time”. I checked the disc count (I alwayscount the smaller number) and said to Iain “31?”. “31” he replied, and went offto tell referee Pete Bhagat the result, while I recorded my loss on my transcriptsheet.

When Pete was announcing the results at the end of the tournament I wasamazed to hear him say that I had won four games, when I thought that I hadwon only three. It turned out that Iain thought that he, not I, had lost our game,and this was what had been recorded as the official result! When I got home Iplayed the game over from my transcript using GTHOR. There must have beena flipping error, because amazingly the outcome of my transcript was 32-32! Sono-one knows what the result really was, and I have therefore decided not to ratethis game either.

To sum up, games whose outcome, whether decided by disc count, timeout,resignation, or penalty, is not in doubt will be rated. If the outcome is in doubt,or if a more appropriate outcome clearly differs from what the referee actuallydecided, then the game is a candidate for being ignored. I hope that this issufficient reassurance that I will not arbitrarily decide not to rate one of yourgames, and I hope that tournament organisers will continue to tell me about anygames whose real outcome was in doubt.

[Editor’s Note. Not everyone agrees with this policy. If you want to write to mewith a dissenting view, please do.]

* * * * * * * * * *

Friendly Rivalry.We are led to believe that Phil Marson and Iain Barrass watch the ratings verycarefully. Phil attained a rating of 1300 before Iain, but Iain has just made it to1400, with Phil still a little way off. Presumably the race is now on for 1500.

47

The Rating List maintained by David Haigh.

This list includes tournaments up to and including the Cambridge Regional, butnot the Manchester Regional. Among those dropping off through inactivity are:Alex Selby, Paul Smith, Helena Verrill, Alec Edgington, Tim Williamson, ColinGraham and Robert Stanton. Some, we hope, will be back.

Meanwhile, an exciting new face has reached the top in NumberOfRat-edGamesPlayed. But who will reach the magic 300 first? In the actual ratinglist, Mike Handel continued his steady rise to the top, Michael Trent leapt in,Iain Barrass went up markedly, and Annemarie Moore acquired status.1 David Shaman 105 1930 32 Lee Evans 32 12352 Imre Leader 295 1831 33 Roy Arnold 274 12083 Graham Brightwell 296 1780 34 David Haigh 291 11944 Neil Stephenson 110 1770 35 Martin Mulvaney 6 11765 Michael Handel 188 1718 36 John Bass 77 11636 Joel Feinstein 250 1685 37 Hamilton Abreu 6 11527 Garry Edmead 105 1671 38 Annemarie Moore 37 11278 Peter Bhagat 288 1651 Jonathan Simpson 12 11279 Guy Plowman 155 1631 40 Margaret Plowman 19 1117

10 Aubrey de Grey 295 1580 41 Colin Hands 32 109411 David Stephenson 126 1541 42 Neil Cuthbertson 47 108212 John Lysons 141 1533 43 Simon Nickson 17 107913 Ian Turner 144 1493 44 Iain Gray 15 107714 Michael Trent 6 1492 45 Graham Chappell 24 107415 Marcus Moore 85 1487 46 Gareth Thomas 18 106816 Jeremy Das 162 1444 47 Maurice Kent 30 106417 Ken Stephenson 147 1438 48 Richard Hemingway 5 105118 Jeremy Rickard 68 1430 49 Ali Turner 81 103919 William Hunter 82 1414 50 Rodney Hammond 46 101720 Iain Barrass 160 1401 51 Adelaide Carpenter 81 101621 Jeremy Benjamin 109 1383 52 Richard Brend 7 101122 Mark Atkinson 75 1345 53 Jim Brewer 57 100423 Phil Marson 153 1334 54 Finton Stephens 7 95524 Phil Brewer 70 1329 55 Liam Stephens 13 89425 Matthew Selby 126 1326 56 Ashley Hammond 26 84926 John Beacock 71 1289 57 Tom Landry 19 83227 Simon John 6 1280 58 Nigel Barforth 6 81328 Mark Wormley 224 1265 59 Steven Verhaegen 6 79229 Martin Fancy 18 1260 60 Gareth Taplin 7 73330 Aiden O’Reilly 12 1245 61 Eileen Forsyth 131 69131 Iain Forsyth 221 1243 62 Packet Of Biscuits 1 639

48